The signal transduction pathways and second messengers which cause a tumor cell to go from a resting to a migrating state are largely unknown. Previous studies with the human melanoma cell line A2058 have shown that the transduction of signals initiating motility differed, depending on the attractant and, in the case of certain extracellular matrix molecules, on whether the attractant was in solution or bound to a substratum. For example, pretreatment-of these cells with pertussis toxin completely abolishes their motility response to their own autocrine motility factor and to type IV collagen in solution, but not to insulin-- like growth factor (IGF-1), or to type IV collagen which is substratum-bound. Sensitivity to PT implicates the involvement of one of the transducer G proteins in these particular biochemical pathways, although its precise identity and mode of action is as yet unknown. Cloning of the gene in A2058 cells which encodes the relevant PT-sensitive G protein should provide answers to some of these questions. An A2058 cDNA library was screened with probes derived from various rat G protein clones. After a secondary screen, over 100 potentially positive clones were identified. The first 25 of these have been carried through a tertiary screen, with 17 still testing positive for one or more classes of G protein (G-alpha-o, G-alpha-i-l, and G-alpha-i-2 have been tentatively identified) . These 17 are being prepared for sequencing and subcloning into antisense expression vectors, in order to identify which class of G protein is actually involved in motility. In addition, mRNA for G-alpha-o and G-alpha-i-2 but not G-alpha-i-l, has been detected in A2058 cells. Western immunoblotting with an antisera specific for G-alpha-o has detected this protein in membranes of A2058 cells; antisera specific for the various classes of G proteins will be used to determine which, if any, class of G proteins redistributes to pseudopodial protrusions during tumor cell migration.